10% Spring Hanger Variability 2

Because stress engineering is not an exact science. Strain critical equipment needs to have better design tolerance. If you use 10% design tolerance, maybe you will actually come within 20%.

The piping and fitting stiffness and weight tolernace used in the load calculations is going to way more than the tolerance of the spring load.

So the point is that as we are not simulating the 100% correct actual conditions (e.g. piping configuration, weights, supports & other item stiffness, surface to surface coefficient of corrosions, real operating temperatures, real ambient temperatures etc.), so to avoid putting the strain sensitive equipment in danger it is better to have less variability (10%) than normal (25%).
@BigInch,
I agree that stress engineering is not an exact science. That is why in COADE Caesar II forum, John Breen once gave his opinion that a system failing with more than 100% stress ratio would not necessarily fail in actual and a system with less than 100% stress ratio is not necessarily safe.
Key is how accurately you can simulate the actual conditions.

You can simulate everything "exactly" similar to what? There is some amount of tolerance in every single variable. For example, can you say for certain that the coefficient of expansion you use for the LACS pipe material will not be 10% more, or less than the typical 0.0000065 in/in-F, that there will be uniform temperature along the entire pipe length, or top to bottom, or that the pipe radius and moment of inertia of the pipe cross section will be exactly uniform for its entire length? One length of pipe may have minimum wall thickness manufacturing tolerance, the adjacent near maximum. It would be impossible to include all potential sources of inaccuracy, even if you had time to do it. That's why we design to our best ability within the time available and use average values to simulate what we know to within reason and set appropriate safety factors to make the result safe. More efficient than 101% accurate simulation. Nobody's going to sit around and wait for a perfectly accurate simulation of a perfectly straight pipe at a perfectly uniform temperature cut and welded to perfect straightness and dimension... The world is not as perfect as a computer simulation can make it appear to be.

BigInch,
Every word you wrote in your post is TRUE!
We have to be within reason and do our best within "available" time to come out with solution.

BI - You forgot the main variable - the pipe fitter with a large wrench / hammer / "jemmy bar"....

Most systems fall down at the point someone has to build it and then all your analysis can be forgotten about.

BI deserves another star for that reply.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

And that "alignment" cable ... attached to a D7

BI/LI,
That's where experience comes in. Where I am now they think that allowing Utilisation factors up to 0.999 is acceptable because it is less than 1.00. When I say that they should be limiting the stress levels to around 80% of the Code they ask why since the Code allows up to 100%. I then tell them that the stress analysis they conduct is not "exact" due to tolerances and also the construction of the pipework will not be exact and the knuckle-rubbing gorillas on site fitting the pipe use turfors/winches or some other mechanical means to align the pipework thereby inducing prestress into the pipework which is not accounted for in the theoretical analysis they tend to start to understand a little.

The responsible design engineer can always decide that the code might not adequately address all, or certain, factors of a specific design at hand and increase the safety factor to whatever level S/HE feels is justified to render the resulting design safe for service. If your boss or client is not pleased with that, tell them that they are free to sign in the little box where it says "ENGR" and then they can do exactly as they want.